bsmschema.models.Node#
- object Node#
A node represents an estimator that applies to a given level of analysis. It contains sufficient information to construct a design matrix, estimate parameter weights (betas) and construct contrasts.
- Config:
extra: str = forbid
- field Level: Literal['Run', 'Session', 'Subject', 'Dataset'] [Required]#
Level of analysis being described.
- field Name: str [Required]#
The name of the node. Must be unique in
BIDSStatsModel.Nodes.
- field GroupBy: List[str] [Required]#
The output statistical maps received from the input node are split along unique combinations of the grouping variables and passed to the model as subsets. If empty, all inputs are passed to a single model to fit. Reserved strings include: “run”, “session”, “subject”, and “contrast”. Any metadata field may be used as a grouping variable.
- field Transformations: Transformations | None#
Specification of transformations to be applied to variables before the construction of the model.
- field Model: Model [Required]#
What model parameters should be included, and how the errors are specified.
- field Contrasts: List[Contrast] | None#
How to linearly weight/combine design matrix columns to generate contrast maps and (optionally) run statistical tests.
- field DummyContrasts: DummyContrasts | None#
A convenient shortcut for specifying contrasts; allows automatic creation of indicator contrasts for either all variables in the design matrix, or all named variables.